基于机器视觉的音叉式表晶谐振裸片分选方法研究

RESEARCH ON THE CLASSIFICATION METHOD OF WATCH CRYSTAL FORK RESONATOR BASED ON MACHINE VISION

针对表面未被银的32768Hz手表用晶体谐振音叉裸片自动检测分选的要求,提出了一种基于机器视觉的晶片自动上料及分选方法并实现了相应的系统,介绍了分选系统的总体结构,着重阐述了视觉分选模块的工作原理,讨论了相应的机器视觉处理算法。实验结果表明该方法能有效地检测音叉片相应外观参数,达到自动分选的目的。

According to the requirements of automatic detection of the crystal resonating tuning fork with 32768Hz which used on watch and its surface is not silvered,a wafer automatic loading and the classification method based on machine vision is presented,and the corresponding system is also achieved.Furthermore,we introduce the overall structure of the classing system,elaborate the working principle of vision classing module.We also discuss the processing algorithms of the machine vision.The result of experiment shows that this method can detect the parameters of the tuning fork＇s appearance effectively,and achieves the goal of automatic selection.